CN103166901B - A kind of 32APSK modulation and demodulation Soft Inform ation computational methods thereof - Google Patents

Abstract

The present invention relates to a kind of 32APSK to modulate and demodulation Soft Inform ation computational methods, belong to the high order modulation-demodulation technical field in wireless digital transmission system; The method comprises: 5 bit informations that modulator inputs are mapped to (I+jQ) baseband complex signal by mapping relations, transmit for rear class.Propose a kind of low complex degree demodulation Soft Inform ation computational methods for this 32APSK modulator approach, these low complex degree demodulation Soft Inform ation computational methods can calculate corresponding 5 bit b from the symbol received ₄b ₃b ₂b ₁b ₀soft Inform ation, for rear end decoding.By demodulation Soft Inform ation computational methods of the present invention, Soft Inform ation computation complexity can be made significantly to reduce, improve high-speed digital transmission hardware realizability, and this simplified calculation method not cause the significantly sacrificing of channel capacity and error performance.

Description

A kind of 32APSK modulation and demodulation Soft Inform ation computational methods thereof

Technical field

The invention belongs to the high order modulation-demodulation technical field in wireless digital transmission system, the calculating of different bit Soft Inform ation in the demodulating process of particularly 32APSK modulation and correspondence.

Background technology

At present, because high-order amplitude-phase shift keying (APSK) technology spectrum efficiency is high, can bandwidth resources be saved, and compared with quadrature amplitude modulation (QAM) technology, lower to the linearity of backend amplifier, thus in bandwidth-limited systems, there is very wide application prospect.In the standard of second generation satellite digital broadcasting (DVB-S2) standard in Europe and " in HRT high-rate telemetry application variable higher level code and modulator approach " (CCSDS131.2-R-1) of consultative committee for space data system (CCSDS), all contain a kind of 32APSK modulation scheme.

The 32APSK demodulation Soft Inform ation computational methods of second generation satellite digital broadcasting (DVB-S2) standard and consultative committee for space data system (CCSDS), comprise the steps.

1) 5 bit information b of modulator input ₄b ₃b ₂b ₁b ₀as shown in table 1 with the mapping relations of (I+jQ) baseband complex signal, wherein R ₁, ρ ₁r ₁, ρ ₂r ₁be respectively the radius of 32APSK constellation point inner ring, middle ring and outer shroud; ρ ₁, ρ ₂in being respectively, the ratio of the radius of outer shroud and inner ring, in table, sequence number illustrates 32 kinds of different mapping relations.

Table 1DVB ?S2 and CCSDS propose 32APSK mapping relations

Mapping relations described by table 1 obtain the 32APSK constellation point diagram shown in Fig. 1, wherein R ₁, ρ ₁r ₁, ρ ₂r ₁be respectively the radius of 32APSK constellation point inner ring, middle ring and most ring;

2) utilize this planisphere to carry out demodulation and obtain Soft Inform ation: by the symbol y=y received _re+ jy _imsubstitute into universal calculation equation (1) and calculate five the Soft Inform ation (LLR (b obtaining and be used for rear end decoding _n)):

$\mathrm{LLR}\left(b_n\right)=\ln \left(\frac{\underset{\mathrm{\γ}\∈{S^n}_1}{\mathrm{\Σ}}{e}^{-\frac{1}{{\mathrm{\σ}}^2}({(y_{\mathrm{re}}-{\mathrm{\γ}}_{\mathrm{re}})}^2+{(y_{\mathrm{im}}-{\mathrm{\γ}}_{\mathrm{im}})}^2)}}{\underset{\mathrm{\γ}\∈{S^n}_0}{\mathrm{\Σ}}{e}^{-\frac{1}{{\mathrm{\σ}}^2}({(y_{\mathrm{re}}-{\mathrm{\γ}}_{\mathrm{re}})}^2+{(y_{\mathrm{im}}-{\mathrm{\γ}}_{i,})}^2)}}\right)---\left(1\right)$

Wherein y=y _re+ jy _imfor the symbol received, S ⁿ ₀, S ⁿ ₁be respectively b _nposition hard decision is the constellation point set of " 0 ", " 1 ", γ=γ _re+ j γ _imfor S ⁿ ₀or S ⁿ ₁in point, σ ²for the variance of additive white Gaussian noise channel.

According to each constellation point 5 bit b ₄b ₃b ₂b ₁b ₀the value that should be judged, is divided into two decision region by planisphere, as shown in Figure 2, dark expression bit hard decision is the region of " 0 ", and light color represents that bit hard decision is the region of " 1 ".As shown in Figure 2, decision region is asymmetric about complex plane reference axis, positive and negative or polar coordinate transform can not judge the region at the symbol place received simply by numerical value.Therefore, said method carry out Soft Inform ation calculate time, γ in formula _re, γ _imvalue is complicated, and Soft Inform ation computing formula comprises a large amount of exponent arithmetic, square operation and division arithmetic, is difficult to reduce computation complexity, is not suitable for hardware implementing high-speed digital transmission.

Summary of the invention

The object of the invention is for overcoming existing 32APSK demodulation Soft Inform ation computing technique computing complexity, amount of calculation is large, be not suitable for a difficult problem for hardware implementing high-speed digital transmission, a kind of novel 32APSK modulation and demodulation Soft Inform ation computational methods thereof are proposed, the Soft Inform ation calculation expression obtained only comprises simple judgement, addition and subtraction, constant and variable multiplying, calculating Soft Inform ation complexity is greatly reduced, is conducive to demodulator hardware and realizes high-speed digital transmission.

A kind of 32APSK modulation that the present invention proposes and demodulation Soft Inform ation computational methods thereof, is characterized in that, comprise the following steps:

1) by 5 bit information b of modulator input ₄b ₃b ₂b ₁b ₀(I+jQ) baseband complex signal is mapped to according to such as table 2 mapping relations, wherein, R ₁, ρ ₁r ₁, ρ ₂r ₁be respectively the radius of 32APSK constellation point inner ring, middle ring and outer shroud, ρ ₁, ρ ₂in being respectively, the ratio of the radius of outer shroud and inner ring, in table, sequence number illustrates 32 kinds of different mapping relations;

Table 2

Under given code efficiency condition, turn to optimization aim, to parameter ρ so that channel capacity is maximum ₁, ρ ₂, R ₁carry out simulation optimization, obtain ρ ₁, ρ ₂, R ₁simulation optimization result is as shown in table 3;

Table 3

2) at receiving terminal, the symbol that demodulator receives is y=y _re+ jy _im, σ ²for awgn channel noise variance, carry out demodulation to the symbol received and export Soft Inform ation, the reduced mechanical model of Soft Inform ation is as follows:

B ₄the expression formula of position (highest order) Soft Inform ation is:

B ₃the expression formula of position Soft Inform ation is:

B ₂the expression formula of position Soft Inform ation is:

$\mathrm{LLR}\left(b_2\right)\&ap;\left\{\begin{array}{c}-\frac{2}{{\mathrm{\&sigma;}}^2}(\frac{2\sqrt{{y^2}_{\mathrm{re}}+{y^2}_{\mathrm{im}}}}{R_1+{\mathrm{\&rho;}}_1{R}_1}-1);(\sqrt{{y^2}_{r3}+{y^2}_{\mathrm{im}}}\&le;\frac{R_1+{\mathrm{\&rho;}}_1{R}_1}{2})\\ -\frac{2}{{\mathrm{\&sigma;}}^2}(\frac{\mathrm{\&pi;}}{3}-|\arctan \frac{y_{\mathrm{im}}}{y_{\mathrm{re}}}\left|\right);(\frac{R_1+{\mathrm{\&rho;}}_1{R}_1}{2}<\sqrt{{y^2}_{\mathrm{re}}+{y^2}_{\mathrm{im}}}\&le;\frac{{\mathrm{\&rho;}}_1{R}_1+{\mathrm{\&rho;}}_2{R}_1}{2})\\ -\frac{2}{{\mathrm{\&sigma;}}^2}(\frac{\mathrm{\&pi;}}{4}-|\arctan \frac{y_{\mathrm{im}}}{y_{\mathrm{re}}}\left|\right);(\frac{{\mathrm{\&rho;}}_1{R}_1+{\mathrm{\&rho;}}_2{R}_1}{2}<\sqrt{{y^2}_{\mathrm{re}}+{y^2}_{\mathrm{im}}})\end{array}\right.$

B ₁the expression formula of position Soft Inform ation is: $\mathrm{LLR}\left(b_1\right)\&ap;-\frac{2}{{\mathrm{\&sigma;}}^2}(1-\frac{2\sqrt{{y^2}_{\mathrm{re}}+{y^2}_{\mathrm{im}}}}{({\mathrm{\&rho;}}_1+{\mathrm{\&rho;}}_2)R_1});$

B ₀the expression formula of position (lowest order) Soft Inform ation is:

Feature of the present invention and beneficial effect:

The 32APSK modulation that the present invention proposes and demodulation Soft Inform ation computational methods thereof, its bit decision region is symmetrical about reference axis, make Soft Inform ation computing formula can obtain approximate efficiently simplification, and propose piecemeal simplified expression for wherein irregular decision region, by the demodulation Soft Inform ation computational methods that the present invention proposes, the computing such as index, variable division is not comprised in Soft Inform ation calculation expression, computation complexity significantly reduces, improve high-speed digital transmission hardware realizability, and this algorithm does not cause the significantly sacrificing of performance by simulating, verifying.

Accompanying drawing explanation

Fig. 1 is the 32APSK constellation point diagram in DVB-S2 standard.

Fig. 2 is the decision region of 32APSK constellation point diagram five bits in DVB-S2 standard.

Fig. 3 is the 32APSK constellation point diagram that the present invention generates.

Fig. 4 is the decision region of 32APSK constellation point diagram five bits that the present invention generates.

Fig. 5 is 32APSK and 32APSK Soft Inform ation BER Simulation comparison diagram of the present invention in DVB-S2 standard.

Fig. 6 is 32APSK and 32APSK channel capacity comparison diagram of the present invention in DVB-S2 standard.

Fig. 7 is a typical digital communication system structured flowchart.

Fig. 8 is the 32APSK modulator approach implementing procedure figure that the present invention proposes.

Fig. 9 is the 32APSK demodulation Soft Inform ation computational methods implementing procedure figure that the present invention proposes.

Embodiment

A kind of 32APSK modulation that the present invention proposes and demodulation Soft Inform ation computational methods thereof, is characterized in that, comprise the following steps:

1) by 5 bit b of modulator input ₄b ₃b ₂b ₁b ₀(I+jQ) baseband complex signal is mapped to, wherein R according to such as table 2 mapping relations ₁, ρ ₁r ₁, ρ ₂r ₁be respectively the radius of 32APSK constellation point inner ring, middle ring and outer shroud, ρ ₁, ρ ₂in being respectively, the ratio of the radius of outer shroud and inner ring, in table, sequence number illustrates 32 kinds of different mapping relations.

Table 2 modulator input-output mappings relation

Make i=1,2,3 represent inner ring, middle ring and outer shroud respectively, take energy normalized as constraints, limit wherein n _ibe the quantity of constellation point on i-th ring, M=32 is constellation point sum, under the condition of given code efficiency, turns to optimization aim, to parameter ρ so that channel capacity is maximum ₁, ρ ₂, R ₁carry out simulation optimization, obtain constellation point diagram parameter ρ under different coding efficiency ₁, ρ ₂, R ₁optimal value; According to ρ ₁, ρ ₂, R ₁optimal value and modulator input generate 32APSK constellation point diagram with the mapping relations of baseband complex signal, the b that each constellation point is corresponding unique ₄b ₃b ₂b ₁b ₀totally 5 bit information, ρ ₁, ρ ₂, R ₁simulation optimization result as shown in table 3.

ρ under table 3 different coding efficiency ₁, ρ ₂, R ₁optimum results

The 32APSK constellation point diagram shown in Fig. 3 is obtained, wherein R according to described mapping relations ₁, ρ ₁r ₁, ρ ₂r ₁be respectively

The radius of 32APSK constellation point inner ring, middle ring and most ring;

According to each constellation point 5 bit ρ ₁, ρ ₂, R ₁each bit value that should be judged, respectively planisphere is divided into five kinds of different decision region: i.e. five kinds of different hard decisions region that to be the region of " bit value is 0 " and hard decision be " bit value is 1 ", as shown in Figure 4, wherein dark expression bit hard decision is the region of " 0 ", and light color represents that bit hard decision is the region of " 1 ".As shown in Figure 4, decision region is symmetrical about complex plane reference axis, positive and negative or polar coordinate transform can judge the region at the symbol place received simply by numerical value.Due to S ⁿ ₀, S ⁿ ₁in constellation point symmetrical about reference axis, therefore γ in formula (1) _re, γ _imequal or opposite number each other, carries out approximate abbreviation by numerical simulation to this formula, obtains simple Soft Inform ation calculation expression:

2) at receiving terminal, it is y=y that demodulator receives symbol _re+ jy _im, σ ²for awgn channel noise variance, carry out demodulation to the symbol received and export Soft Inform ation, the reduced mechanical model of Soft Inform ation is as follows:

B ₄the expression formula of position (highest order) Soft Inform ation is:

_bthe expression formula of 3 Soft Inform ation is:

B ₂the expression formula of position Soft Inform ation is:

$\mathrm{LLR}\left(b_2\right)\&ap;\left\{\begin{array}{c}-\frac{2}{{\mathrm{\&sigma;}}^2}(\frac{2\sqrt{{y^2}_{\mathrm{re}}+{y^2}_{\mathrm{im}}}}{R_1+{\mathrm{\&rho;}}_1{R}_1}-1);(\sqrt{{y^2}_{r3}+{y^2}_{\mathrm{im}}}\&le;\frac{R_1+{\mathrm{\&rho;}}_1{R}_1}{2})\\ -\frac{2}{{\mathrm{\&sigma;}}^2}(\frac{\mathrm{\&pi;}}{3}-|\arctan \frac{y_{\mathrm{im}}}{y_{\mathrm{re}}}\left|\right);(\frac{R_1+{\mathrm{\&rho;}}_1{R}_1}{2}<\sqrt{{y^2}_{\mathrm{re}}+{y^2}_{\mathrm{im}}}\&le;\frac{{\mathrm{\&rho;}}_1{R}_1+{\mathrm{\&rho;}}_2{R}_1}{2})\\ -\frac{2}{{\mathrm{\&sigma;}}^2}(\frac{\mathrm{\&pi;}}{4}-|\arctan \frac{y_{\mathrm{im}}}{y_{\mathrm{re}}}\left|\right);(\frac{{\mathrm{\&rho;}}_1{R}_1+{\mathrm{\&rho;}}_2{R}_1}{2}<\sqrt{{y^2}_{\mathrm{re}}+{y^2}_{\mathrm{im}}})\end{array}\right.$

B ₁the expression formula of position Soft Inform ation is: $\mathrm{LLR}\left(b_1\right)\&ap;-\frac{2}{{\mathrm{\&sigma;}}^2}(1-\frac{2\sqrt{{y^2}_{\mathrm{re}}+{y^2}_{\mathrm{im}}}}{({\mathrm{\&rho;}}_1+{\mathrm{\&rho;}}_2)R_1});$

B ₀the expression formula of position (lowest order) Soft Inform ation is:

Modulate and demodulation Soft Inform ation computational methods for the 32APSK modulation-demo-demodulation method proposed in DVB-S2 and this 32APSK, introduce error correcting code coding and decoding simulated environment simultaneously, carry out emulation testing, obtain the error-correcting performance of Soft Inform ation in two schemes respectively, result as shown in Figure 5.As seen from Figure 5, two ber curves overlap substantially, the 32APSK modulation that the present invention proposes and demodulation Soft Inform ation computational methods thereof are described, do not cause the significantly sacrificing of error performance.

Adopt awgn channel calculation of capacity formula, in contrast DVB-S2 standard, the channel capacity of the 32APSK constellation point diagram that 32APSK and this method generate emulates, and its channel capacity is with bit signal to noise ratio (E _b/ N ₀) change as shown in Figure 6.Both curve co-insides, illustrate that channel capacity does not have obvious decline as seen from Figure 6.

Below in conjunction with accompanying drawing and embodiment, the present invention is further described, and the present invention is applied in common 32APSK modulation /demodulation digital communication system, is described below for a simple digital communication system.The structured flowchart of this communication system as shown in Figure 7, information source produces bit stream input error correcting encoder, the bit stream every 5 one group that encoder exports sends into this 32APSK modulator, the mapping method of this 32APSK modulator by describing in table 2, produce the symbol of (I+jQ) complex space, be modulated to radio frequency band by radio-frequency modulator.

Radiofrequency signal can introduce noise after awgn channel, and the symbol that radio-frequency (RF) demodulator demodulates comprises noisy symbol.32APSK demodulator can judge the cut zone belonging to this symbol, substitutes into corresponding reduced mechanical model, calculates b ₀b ₁b ₂b ₃b ₄soft Inform ation corresponding respectively, send into decoder and carry out decoding, decode results is delivered to the stay of two nights, completes communication process.

For 32APSK baseband-modem module in Fig. 7, detailed modulation implementing procedure as shown in Figure 8.For demodulation Soft Inform ation computing module in Fig. 7, detailed implementing procedure as shown in Figure 9.

Below using the bit information of an actual information source generation as embodiment, illustrate that method of the present invention is as follows:

1) suppose that encoder encodes efficiency is 4/5,5 bits of modulator input are 10101;

2) according to the optimum results ρ of parameter ₁=2.72, ρ ₂=4.87, R ₁=0.260, and the mapping relations of table 2, the output symbol obtaining modulator is y=-0.184+j0.184.

3) if modulator continues to input new bit, so repeat 2) in process calculate.

An actual symbol for a demodulator input illustrates demodulating process:

What suppose that demodulator inputs is y=-0.45+j0.7794, then y containing noisy baseband signalling _re=-0.45, y _im=0.7794.

The expression formula provided by the present invention of above-mentioned reception Information Pull calculates the Soft Inform ation of each bit in 5 bits:

$\mathrm{LLR}\left(b_4\right)\&ap;-\frac{2}{{\mathrm{\&sigma;}}^2}{y}_{\mathrm{re}}=\frac{0.9}{{\mathrm{\&sigma;}}^2}$

$\mathrm{LLR}\left(b_3\right)\&ap;-\frac{2}{{\mathrm{\&sigma;}}^2}{y}_{\mathrm{im}}=-\frac{1.5588}{{\mathrm{\&sigma;}}^2}$

Because $\sqrt{{y^2}_{\mathrm{re}}+{y^2}_{\mathrm{im}}}=0.900$ Meet $(\frac{R_1+{\mathrm{\&rho;}}_1{R}_1}{2}<\sqrt{{y^2}_{\mathrm{re}}+{y^2}_{\mathrm{im}}}\&le;\frac{{\mathrm{\&rho;}}_1{R}_1+{\mathrm{\&rho;}}_2{R}_1}{2}),$ Therefore

$\mathrm{LLR}\left(b_2\right)\&ap;-\frac{2}{{\mathrm{\&sigma;}}^2}(\frac{\mathrm{\&pi;}}{3}-|\arctan \frac{y_{\mathrm{im}}}{y_{\mathrm{re}}}\left|\right)=\frac{\mathrm{\&pi;}}{{3\mathrm{\&sigma;}}^2}$

$\mathrm{LLR}\left(b_1\right)\&ap;-\frac{2}{{\mathrm{\&sigma;}}^2}(\frac{{\mathrm{\&rho;}}_1{R}_1+{\mathrm{\&rho;}}_2{R}_1}{2}-\sqrt{{y^2}_{\mathrm{re}}+{y^2}_{\mathrm{im}}})=-\frac{0.1734}{{\mathrm{\&sigma;}}^2}$

Because $\sqrt{{y^2}_{\mathrm{re}}+{y^2}_{\mathrm{im}}}=0.900$ Meet $(\frac{R_1+{\mathrm{\&rho;}}_1{R}_1}{2}<\sqrt{{y^2}_{\mathrm{re}}+{y^2}_{\mathrm{im}}}\&le;\frac{{\mathrm{\&rho;}}_1{R}_1+{\mathrm{\&rho;}}_2{R}_1}{2}),$ Therefore

$\mathrm{LLR}\left(b_0\right)\&ap;-\frac{2}{{\mathrm{\&sigma;}}^2}\left(\right|\arctan \frac{y_{\mathrm{im}}}{y_{\mathrm{re}}}|-\frac{\mathrm{\&pi;}}{6})=-\frac{1}{{\mathrm{\&sigma;}}^2}.$

Corresponding 5 bit b can be calculated thus ₀b ₁b ₂b ₃b ₄the Soft Inform ation of each, decoder Soft Inform ation being used for rear end carries out decoding judgement, completes the function of modulation /demodulation.And the computation complexity of Soft Inform ation reduces greatly.If demodulator receives new symbol, then repeat above-mentioned steps.

Claims (1)

1. 32APSK modulation and demodulation Soft Inform ation computational methods thereof, is characterized in that, comprise the following steps:

1) by 5 bit information b of modulator input ₄b ₃b ₂b ₁b ₀i+jQ baseband complex signal is mapped to according to such as table 2 mapping relations, wherein, R ₁, ρ ₁r ₁, ρ ₂r ₁be respectively the radius of 32APSK constellation point inner ring, middle ring and outer shroud, ρ ₁, ρ ₂in being respectively, the ratio of the radius of outer shroud and inner ring, in table 2, sequence number illustrates 32 kinds of different mapping relations;

Wherein table 2 is:

Under given code efficiency condition, turn to optimization aim, to parameter ρ so that channel capacity is maximum ₁, ρ ₂, R ₁carry out simulation optimization, obtain ρ ₁, ρ ₂, R ₁simulation optimization result is as shown in table 3;

Wherein table 3 is:

2) at receiving terminal, the symbol that demodulator receives is y=y _re+ jy _im, σ ²for awgn channel noise variance, carry out demodulation to the symbol received and export Soft Inform ation, the reduced mechanical model of Soft Inform ation is as follows:

B ₄the expression formula of position Soft Inform ation is:

B ₃the expression formula of position Soft Inform ation is:

B ₂the expression formula of position Soft Inform ation is:

B ₁the expression formula of position Soft Inform ation is:

B ₀the expression formula of position Soft Inform ation is:

。